A content QA audit scores a page across five dimensions that determine whether it ranks, earns trust, and gets cited by AI search platforms. This checklist distills the audit system we use at xeo.works into 50 verifiable checks any B2B SaaS team can run on their own content.

How to use this checklist: Audit one page at a time. Score each check as PASS (2 points), PARTIAL (1 point), or FAIL (0 points). Sum the scores across all 50 checks for a raw total out of 100. Pages scoring below 55 are actively hurting your domain authority.

40%

Reduction in low-quality content Google targeted in its March 2024 Core Update

Search Engine Journal

90%

Of the buyer journey completed before contacting a vendor

LinkedIn / Demand Gen Report

20-50%

Traffic decline for brands not optimized for AI search

McKinsey 2025

Dimension 1: Technical SEO (10 Checks)

Technical SEO is the foundation layer. If your page has broken meta tags, missing schema, or a mangled heading hierarchy, nothing else matters — search engines can't properly index what they can't parse. These 10 checks verify that the on-page fundamentals are correctly implemented before you worry about content quality.

Technical SEO

Are the on-page SEO fundamentals correctly implemented?

Single H1 containing primary keyword

Exactly one H1 tag on the page, and it includes the primary keyword you're targeting. Multiple H1s confuse search engines about the page's topic. Check your CMS or HTML output — some templates accidentally generate two.

Primary keyword in first 100 words

Your target keyword appears naturally within the first 100 words of body content. This signals topical relevance immediately. Count from the start of prose — not from the nav or header.

Meta title under 60 characters with keyword

The title tag is 60 characters or fewer, includes the primary keyword, and follows a consistent format (e.g., ‘Primary Keyword | Brand’). Titles over 60 chars get truncated in SERPs.

Meta description under 155 characters with keyword and value prop

The meta description is 155 characters or fewer, includes the primary keyword, and gives users a reason to click. A description without a value proposition is a missed conversion opportunity.

Self-referencing canonical URL

The page has a canonical tag pointing to its own URL (not a different page). Use the full absolute URL. Mismatched canonicals cause indexing confusion and can consolidate your ranking signals onto the wrong page.

OpenGraph and Twitter card metadata

Both OG tags (og:title, og:description, og:image) and Twitter card meta are present. These control how your page appears when shared on social platforms and in some AI system previews.

Correct schema types for page type

The page includes the right JSON-LD schema types: Service + BreadcrumbList + FAQ for service pages, Article + BreadcrumbList for blog posts, DefinedTerm for glossary pages. Wrong schema types confuse search engines about page purpose.

No schema placeholders or TODOs

Search your structured data for PLACEHOLDER, TODO, TBD, or template brackets. Any placeholder in live schema is worse than having no schema — it tells search engines you shipped unfinished work.

Valid heading hierarchy (H1 > H2 > H3)

Headings follow a clean hierarchy with no skipped levels. An H3 should always be nested under an H2, never directly under the H1. Screen readers and search engines both rely on this structure to understand content relationships.

All images have descriptive alt text

Every image has an alt attribute that describes what the image shows — not the filename, not ‘image1.png.’ Alt text serves accessibility, helps search engines understand visual content, and provides context to AI systems.

Dimension 2: Internal Linking & Structure (10 Checks)

Internal links are how search engines discover and weight your pages. A page with five relevant internal links sends a stronger topical signal than one floating in isolation. These checks also cover content structure — word count, paragraph length, and the presence of structured elements that make long-form content scannable.

Hub Pages

Core topic pages that concentrate authority

Spoke Pages

Supporting content linking back to hubs

Cross-Links

Related pages referencing each other

CTA Links

Conversion paths from content to contact

Internal Linking & Structure

Does the page connect to your site architecture and meet structural standards?

Minimum internal links met

Landing and service pages need at least 5 internal links to other pages on your site. Blog posts and glossary pages need at least 3. Count actual navigation-independent links in the body content — not nav or footer links.

Links to hub/pillar pages present

Every page should link to at least one hub or pillar page in your topic cluster. If you have a main ‘SEO for B2B SaaS’ page, your supporting content should link to it. This is how you build topical authority.

Hub link appears within first 300 words

At least one link to a hub or pillar page appears in the opening section of the content. Early links carry more weight and help search engines immediately understand how this page fits your site's topic architecture.

Anchor text varies across links

Links to the same destination use different anchor text. Repeating the exact same anchor phrase for every link to a page looks manipulative. Use natural variations that describe different aspects of the destination.

Cross-links to related pages

The page links to at least 2 related pages of the same type. Vertical pages should cross-link to other verticals. Service pages should reference related services. This creates the web of relevance search engines reward.

No broken internal links

Every internal link resolves to a live page. Broken links waste crawl budget and create dead ends for users. Run a link checker or manually verify that each href points to an existing page.

CTA links to a conversion page

At least one call-to-action button or link points to your contact or demo page. Content without a conversion path is a missed opportunity. Place CTAs after demonstrating value — not before.

Word count meets minimum for page type

Landing pages: 2,000+ words. Blog posts: 1,500+ words (pillar posts: 3,000+). Glossary pages: 500+ words. Thin content signals low authority. But padding to hit a number is worse — every sentence should earn its place.

Paragraphs are 4 sentences or fewer

Scan all paragraphs. None should exceed 4 sentences. Long paragraphs kill readability on screens and signal AI-generated content (which tends toward dense blocks). Break them up.

At least 1 structured element per 1,000 words

Tables, lists, comparison grids, stat cards, or visual diagrams should appear at least once per 1,000 words of body content. All-prose pages are harder to scan and less likely to earn featured snippets or AI citations.

Dimension 3: E-E-A-T & Authority (10 Checks)

E-E-A-T (Experience, Expertise, Authoritativeness, Trustworthiness) isn't a direct ranking factor — it's the framework Google's quality raters use to evaluate content. But the signals that demonstrate E-E-A-T are the same signals that make content persuasive to buyers and citable by AI systems. These checks verify your content reads like it was written by a practitioner, not assembled from the top 10 search results.

Reads Like a Textbook

“Many B2B SaaS companies find that implementing a comprehensive SEO strategy can lead to significant improvements in organic traffic and lead generation over time.”

Reads Like a Practitioner

“When we audit B2B SaaS sites, the same pattern shows up: 40+ blog posts, zero attributable pipeline. The content exists, but it targets informational queries that never convert.”

E-E-A-T & Authority

Does the content demonstrate real practitioner experience and domain authority?

Experience markers from direct practice

Content contains at least 3 statements showing firsthand experience: ‘we audit,’ ‘we've tested,’ ‘when we work with,’ ‘in our experience.’ These distinguish practitioner knowledge from aggregated information.

Named methodology or framework

The page references at least one specific, named methodology unique to your organization. Generic advice (‘create quality content’) is interchangeable. A named framework (‘our 5-Step AEO Framework’) demonstrates original thinking.

Specific data points with sources

At least 3 verified data points with attributed sources. ‘Studies show’ without a citation fails. ‘According to Gartner's 2026 survey, 38% of software buyers start with AI chatbots’ passes. Specificity builds trust.

Honest limitations stated

The page includes at least one boundary or limitation: ‘this approach won't work if...,’ ‘we don't guarantee results,’ ‘this is specifically for X, not Y.’ Content that claims to solve everything earns trust from no one.

Specificity over generality throughout

Content uses concrete numbers, names, and examples instead of vague claims. Flag phrases like ‘many companies,’ ‘significant growth,’ or ‘various industries’ — replace with specific figures, named examples, and precise descriptions.

At least one contrarian or differentiated take

The page takes at least one clear stance that distinguishes it from generic advice on the same topic. If 10 competitors say the same thing, saying it an 11th time adds no authority. State what you believe differently — and back it with evidence.

FAQ answers are self-contained

Each FAQ answer works as a standalone response without needing the question for context. ‘Yes, it does’ fails. A complete statement that could be extracted and cited on its own passes.

Explicit ‘not a fit’ or scope acknowledgment

Service and landing pages state who the offering is NOT for. Blog posts acknowledge the limits of their advice. This honesty signal is rare — and both human readers and AI systems reward it.

No unresolved placeholders

Zero instances of [PLACEHOLDER], [TODO], [NEEDS VERIFICATION], [CASE STUDY], or TBD in published content. Every placeholder is a trust signal in reverse — it tells readers (and AI systems) the content isn't finished.

Fair competitor references

When competitors are mentioned, they're referenced factually with strengths noted before limitations. No disparagement, no vague ‘some agencies’ when you mean a specific competitor. Name them, be fair, and let the reader decide.

Dimension 4: Anti-Slop Detection (10 Checks)

“Slop” is AI-generated filler that sounds professional but says nothing. It's the phrases, structures, and patterns that make content feel like it was assembled by a language model rather than written by someone who does the work. Google's March 2024 Core Update targeted a 40% reduction in this kind of content. These checks help you catch it before it erodes your domain authority.

Slop Detection Layers

Generic Benefit Statements

Grow your business, save time and money, take X to the next level

Vague Authority Claims

Studies show, research indicates, experts agree — without any citation

Structural Monotony

Every section follows the same format with identical paragraph cadences

Banned Phrases

Specific words and phrases that signal AI-generated filler (leverage, synergy, cutting-edge)

Anti-Slop Detection

Does the content read like it was written by a practitioner or assembled by a language model?

Zero banned phrases

Search the page for: leverage, best-in-class, synergy, cutting-edge, state-of-the-art, solutions (standalone), unlock, unleash, game-changer, disruptive, thought leadership (self-referential), low-hanging fruit, move the needle, at the end of the day, in today's digital landscape, one-stop shop, guaranteed results. Any occurrence is a fail.

Consistent voice (no ‘I/my’ on team pages)

If your content uses ‘we’ for the organization, it should be consistent throughout. Switching between ‘I’ and ‘we’ mid-page is a common AI-generation artifact. Exception: personal blog posts or founder-attributed content.

No filler transitions

Search for: ‘Let's dive in,’ ‘Without further ado,’ ‘In this section we'll explore,’ ‘Let's get started,’ ‘Now let's talk about,’ ‘Moving on to.’ These are padding sentences that add zero information. Cut them.

No vague authority claims

Search for: ‘studies show,’ ‘research indicates,’ ‘experts agree,’ ‘it's widely known,’ ‘it's well-established’ — all without a specific citation following. Either name the source or rewrite the sentence without the appeal to unnamed authority.

No adjective stacking

Search for 3 or more consecutive adjectives before a noun: ‘comprehensive, innovative, cutting-edge solution.’ This is a hallmark of AI-generated copy. One well-chosen adjective is stronger than three generic ones.

No generic benefit statements

Flag: ‘save time and money,’ ‘grow your business,’ ‘take your X to the next level,’ ‘streamline your operations,’ ‘maximize your potential,’ ‘drive meaningful results.’ Replace with specific outcomes tied to your actual service or product.

Section openers are specific

Read the first sentence of every H2 section. Each should state a specific claim, data point, or question — not generic setup like ‘In this section, we'll cover...’ or ‘It's important to understand...’ Specific openers earn attention. Generic ones lose it.

Prose rhythm and structure vary across sections

Content should alternate between prose, lists, tables, callouts, and visual elements. If every section follows the same structure (intro paragraph, 3 bullet points, closing paragraph), the content feels templated — because it probably is.

No fabricated data

Zero invented statistics, case studies, testimonials, or client names. Every data point traces back to a verifiable source. A single fabricated stat discovered by a reader or AI system can destroy months of authority building.

Practitioner knowledge test

The page contains at least 3 elements requiring domain-specific knowledge: specific tool workflows, industry regulations by name, named competitor analysis, attribution methodology details, or vertical-specific buyer behavior. If ChatGPT could produce identical content from a generic prompt, the page fails.

Dimension 5: Content Structure for AI Extraction (10 Checks)

AI search platforms — ChatGPT, Perplexity, Claude, Google AI Overviews — don't read pages the way humans do. They extract individual passages, definitions, and structured data. Content optimized for extraction gets cited. Unstructured prose gets skipped. These checks verify your content is shaped for the way AI systems actually select what to cite.

Content Structure for AI Extraction

Can AI systems extract clean, citable passages from this page?

Direct answer or TLDR in first 300 words

The first 300 words contain a clear, direct answer to the page's core search query. For long-form pages (2,000+ words), include a visually distinct Quick Answer or TLDR block under 60 words. If the AI only reads your opening, does it get a usable answer?

Self-contained first sentence in each H2 section

Every major section opens with a 40-60 word sentence that works as a standalone answer. Test: cover the heading and read only the first sentence. Does it state a specific claim an AI could extract and cite? If it says ‘This section covers...’ or ‘As mentioned above,’ it fails.

Definition blocks use ‘X is Y’ format

When defining a concept, lead with a direct declaration: ‘Schema markup is structured data vocabulary...’ not ‘When it comes to structured data...’ AI systems extract ‘X is Y’ patterns as complete citation blocks.

Numbered frameworks with bold step labels

Multi-step processes use numbered lists with bold step names and one-sentence definitions per step. AI systems cite numbered frameworks as complete blocks. Processes described only in prose rarely get extracted.

Comparison tables for any X vs Y content

Any comparison content uses HTML tables with clear column headers and at least 3 rows. Tabular data gets extracted whole by AI platforms. Prose comparisons (‘Option A is better than Option B because...’) are harder to cite accurately.

Section headings are descriptive, not clever

H2 and H3 headings describe their content plainly. ‘How Schema Markup Improves AI Citations’ works. ‘The Schema Secret’ doesn't. AI systems use headings to index content — clever headings make your content harder to find.

No vague opening paragraph

The page's first paragraph starts with substance. Not ‘In today's landscape...’ or ‘As businesses increasingly...’ Lead with a specific claim, definition, or data point within the first 2 sentences. Context-setting openings get skipped by AI extractors.

Lists use semantic HTML elements

List content uses actual <ul> or <ol> elements, not dashes or asterisks formatted as prose. Structured lists get extracted cleanly by AI systems. Pseudo-lists embedded in paragraphs don't.

Key data points in structured format

Important statistics live in tables, callout cards, or list items — not buried in the middle of a paragraph. AI systems extract structured data more reliably than inline numbers. If a stat matters, give it visual prominence.

FAQ answers start with a direct response

Each FAQ answer begins with a direct, self-contained response in the first sentence. No ‘It depends,’ no ‘Great question!’ — lead with the answer. The response must work if extracted without the question text.

Scoring Guide

Score each of the 50 checks as PASS (2 points), PARTIAL (1 point), or FAIL (0 points). Sum for a raw total out of 100.

Score	Verdict	Action
90-100	Ship it	Production ready. Strong across all 5 dimensions. Monitor and iterate.
75-89	Minor fixes	Flag specific failed checks and fix before deploying. Most pages land here on first audit.
55-74	Significant gaps	Fails E-E-A-T, anti-slop, or AI extraction structure. Major revision needed before this page helps your domain.
Below 55	Full rewrite	Generic content that actively hurts domain authority and is invisible to AI search. Start over with practitioner input.

Minimum scores by page type: Homepage: 85+. Blog posts: 80+. Landing and service pages: 75+. Glossary and reference pages: 65+.

Priority order for fixes: Start with Dimension 4 (Anti-Slop) and Dimension 3 (E-E-A-T) — these have the highest impact on content quality perception by both Google and AI systems. Technical SEO issues (Dimension 1) are easier to fix but matter less if the content itself reads like template output.

What to Do Next

Run this checklist on your highest-traffic page first. If it scores below 75, your best-performing content has structural weaknesses that are costing you rankings, trust, and AI citations.

For the AI extraction dimension specifically, see the AEO Citation Readiness Checklist — a deeper 43-check audit focused entirely on getting cited by ChatGPT, Perplexity, Claude, and Google AI Overviews.

For the full SEO methodology for B2B SaaS behind these checks, including how technical SEO, content authority, and AI optimization work together, read the service overview. For hands-on help running this audit across your site, get in touch. We start every engagement with a content audit — and the results are usually eye-opening.